Although first-generation adenovirus vectors have many attractive features for gene therapy applications, they do not persist in vivo due to the strong host immune response against adenovirus-encoded proteins. The development of so-called "gutless" or fully-deleted adenovirus vectors (FD-AdV) which do not encode any adenovirus genes, allowed this limitation to be at least partially overcome. However the inability of these vectors to replicate in the absence of a helper virus complicates the vector propagation and purification procedures and results in the contamination of all FD-AdV stocks by the helper virus, which is potentially as immunogenic as first-generation vectors. We propose FD-AdVs can be safely and efficiently generated and propagated through the substitution of the infectious adenovirus helper by a baculovirus/adenovirus (Bac/Ad) hybrid virus that carries a Cre recombinase-excisable copy of an adenovirus genome that is deleted for El and the packaging signal (helper genome). Additionally, the adenovirus ITRs in this construct will be joined to form an ITR junction known to be a functional adenovirus origin of replication, so that Cre-mediated excision will generate a replication-competent, packaging-deficient circular adenovirus genome. Such a hybrid virus will be propagated in insect cells and used to infect Cre-expressing, El-complementing cells in which FD-AdVs are produced. In these cells, the circular adenovirus genome will be excised from the hybrid. Adenovirus genes expressed from this helper genome will complement the replication of both the FD-AdV and the helper genome, while only the FD-AdV will be packaged into virions. Unlike the helper virus in the traditional helper-dependent system, the Bac/Ad hybrid helper cannot reproduce in mammalian cells and the excised helper genome cannot be incorporated into virus particles due to the lack of the adenovirus packaging signal. To prevent generation of packaging-competent adenovirus through homologous recombination, two approaches will be applied. In the first, the size of the helper genome will be increased to exceed the adenovirus packaging limit through the insertion of a staffer fragment into the E3 region of the helper genome. In the second approach, the FD-AdV will be propagated in the cell lines that are not prone to the generation of replication-competent adenovirus (RCA), This will allow the production of FD-AdVs that are free of contamination by both helper virus and RCk Additionally, a similar system will be developed to produce FD-AdVs based on unrelated adenovirus serotypes that potentially can overcome the problems associated with pre-existing immunity and allow repeat administration of FD-AdVs based on different serotypes. Once these systems are optimized, they will be used for production of safer, more persistent FD-AdVs for gene therapy applications.